NIH funded studies of pediatric

We formatted and analyzed all NIH funded registered studies of pediatric patients. We applied standardization and normalization methods using Elsevier’s Merged Medical Taxonomy terms and the corresponding ICD 10 codes, EMBASE taxonomy and natural language processing methods. We performed systemic appraisal of completed studies according to treatment allocation, masking of treatment status, early termination, and exclusion enrolled subjects from the analyses. We calculated treatment effect for all posted hypotheses and provided the interpretation of treatment effects according to statistical significance, direction, and magnitude of benefits and harms.

本研究对所有由美国国立卫生研究院（National Institutes of Health，NIH）资助的儿科患者注册研究进行了格式化处理与分析。我们采用爱思唯尔（Elsevier）合并医学分类系统术语、对应国际疾病分类第10版（International Classification of Diseases 10th Revision, ICD-10）编码、荷兰医学文摘数据库（Excerpta Medica database, EMBASE）分类体系，结合自然语言处理（Natural Language Processing, NLP）方法完成了数据标准化与归一化操作。我们依据治疗分配方案、治疗状态设盲情况、研究提前终止情形以及分析中排除入组受试者的规则，对已完成的研究开展系统性评估。针对所有已发布的研究假设，我们计算了治疗效应，并依据统计学显著性、效应方向以及获益与损害的程度，对治疗效应进行了解释说明。